Issue with "Expression Error" in Hypermesh NVH2024 Correlation - Optimization
Hello,
I am using the Correlation - Optimization function in Hypermesh NVH2024 (the successor to NVH Director) and encountered an "Expression Error" that I would like to resolve.
To accurately model the FEM system, I am comparing the Modal Assurance Criterion (MAC) between experimental modal analysis results and simulation results in order to update the simulation model.
For modal testing, I used Simcenter software and hardware to generate a .unv file. The simulation files were created using Hypermesh 2024, generating .h3d and .fem files.
Following standard guidelines, I set the .unv file as Reference Data and the .h3d file as Correlation Data and successfully compared the MAC values. I then attempted to optimize the PBUSH stiffness parameters (K1–K3) to update the simulation model.
After properly defining the design variables, objective, and constraints, I ran the [Evaluate]. Once the analysis was completed, I encountered the following issue:
- The Initial MAC and Optimized MAC results display EXPRESSION_ERROR instead of values.
What could be the cause of this error?
It’s not a major issue, but I was wondering about something. When I reach the Job Submission step and press the "Submit Job" button, nothing appears to happen within the Hypermesh NVH interface. However, in Windows Task Manager, I can see that processes such as Tclsh Application, Altair HyperStudy Batch, optistruct_2024_win64.exe, and radflex_2024_win64.exe are running, indicating that the analysis is in progress.
However, since no separate window or interface appears to indicate the progress of the simulation, I am unable to confirm whether the analysis is truly running or how much progress has been made.
Is this the expected behavior of the software?
Lastly, when I tried to resolve this issue using HyperStudy instead of Hypermesh NVH, I encountered the same Expression Error.
I suspect that the issue might be due to the software failing to load variables from the .unv file, but I am not entirely sure.
Thank you in advance for your support.
Answers
-
Hi. I don't have a clear hypothesis on what might be happening unfortunately. But i can suggest one thing that I would do (if you haven't done it yet).
You could repeat the process with a simple, public model and replace the reference unv with an h3d of a similar simple model, maybe with different material of a similar modification. Depending on the result of this test, you might be able to find a clue, but also you could share the file with the community and another set of eyes might find /come up with something.
My two cents.
0